Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2023 Jan 19.
Published in final edited form as: Int J Eat Disord. 2020 Nov 13;54(4):527–534. doi: 10.1002/eat.23407

The association between leptin and weight maintenance outcome in anorexia nervosa

Youngjung Kim 1, Jonathan Hersch 2, Lindsay P Bodell 3, Janet Schebendach 4, Tom Hildebrandt 1, B Timothy Walsh 4, Laurel E S Mayer 4
PMCID: PMC9851598  NIHMSID: NIHMS1690573  PMID: 33185933

Abstract

Objective:

Relapse after weight restoration in anorexia nervosa (AN) is a critical problem. Higher body fat percentage after weight gain has been shown to predict better weight maintenance outcome. Leptin, a fat-derived hormone, has been associated with progress during weight gain, but its association with weight maintenance is unknown. This study aims to determine whether leptin levels after weight restoration in AN are associated with weight maintenance.

Method:

Participants were 41 women with AN hospitalized for inpatient treatment. Participants were evaluated 2–4 weeks after weight restoration to body mass index (BMI) ≥ 19.5 kg/m2 for plasma leptin and body composition. Weight maintenance outcome was defined by whether a participant maintained a BMI of at least 18.5 kg/m2 at the end of 1 year following hospital discharge.

Results:

Twenty (48.8%) out of 41 patients maintained their weight at 1 year. Percent body fat and leptin were significantly higher in the group who maintained weight (body fat, p = .004, Hedges’ g = 0.944; log-leptin, p = .010, Hedges’ g = 0.821), but there were no differences in predischarge BMI, duration of illness, and duration of amenorrhea. Using regression modeling, only higher log-leptin (pWald = .021) and percent body fat (pWald = .010), as well as fat-adjusted leptin (pWald = .029), independently predicted weight maintenance at 1 year.

Discussions:

Our findings suggest that for acutely-weight restored women with AN, higher predischarge leptin measurements are associated with better outcome in the year following treatment. Prospective studies examining leptin as well as other parameters of metabolic health could offer insights into biomarkers that may improve clinical outcomes.

Keywords: anorexia nervosa, body composition, leptin, metabolism, treatment outcome

1 |. INTRODUCTION

Anorexia nervosa (AN) is characterized by a relentless pursuit of thinness through starvation together with high mortality rates despite over a century of search for effective treatments (Arcelus, Mitchell, Wales, & Nielsen, 2011). The psychopathology of AN—extreme fear of fat and eating—poses an inherent challenge in treatment aimed at weight gain which involves regular eating. Even after successful treatment through an intensive treatment program, relapse rates can be above 50% and patients often have a chronic course of illness with relapsing–remitting periods (Khalsa, Portnoff, McCurdy-McKinnon, & Feusner, 2017; Pike, 1998; Steinglass & Walsh, 2016). In a 25-year follow-up, over 70% of the 92% living patients had relapsed to some form of eating disorder (Fichter, Quadflieg, Crosby, & Koch, 2017). Considering such high rates of relapse and associated morbidity and mortality, robust biomarkers of outcome could be helpful in identifying individuals with higher risk of relapse who may benefit from more intensive relapse prevention measures.

Studies have pointed to psychological measures such as higher body image disturbances (Keel, Dorer, Franko, Jackson, & Herzog, 2005), higher weight concern (Carter, Blackmore, Sutandar-Pinnock, & Woodside, 2004), and lower motivation to recover during and after treatment (Carter et al., 2012), as potential risk factors for relapse (reviewed in Berends, Boonstra, & van Elburg, 2018). Among the biological parameters, older age (Berends et al., 2016), lower body mass index (BMI) (Aguera et al., 2015; El Ghoch, Calugi, Chignola, Bazzani, & Dalle Grave, 2016; Kaplan et al., 2009), lower body fat percentage (Bodell & Mayer, 2011; Mayer et al., 2007), measured after weight restoration have been associated with less favorable maintenance outcome in adult patients with AN. Among them, evidence suggests that attaining a higher body fat percentage during weight restoration treatment may be protective against relapse during the maintenance phase. However, repeatedly measuring body fat percentage can be challenging, and a less costly surrogate biomarker is desirable. We hypothesized that one such potential surrogate of body fat percentage may be an adipocyte-derived hormone leptin that can be easily measured from blood.

Leptin regulates energy homeostasis and food intake across the weight spectrum, and its concentration in blood is highly correlated with the total amount of body fat, reflecting both long- and short-term energy status (Considine et al., 1996; Zhang et al., 1994). In addition to energy homeostasis, reduced leptin in AN is associated with amenorrhea, and menstrual return accompanies the long-term recovery from AN (Mantzoros et al., 2011; Misra & Klibanski, 2014), supporting a critical role of leptin in endocrine health (Chou et al., 2011). Genome-wide association studies have identified risk loci for AN that implicates dysregulated leptin signaling (Li, et al., 2017), and because underweight patients with AN have significant reductions in body fat and leptin (Focker et al., 2011; Grinspoon et al., 1996; Hebebrand et al., 1995; Lear, Pauly, & Birmingham, 1999), leptin has been suggested as a potential biomarker for screening of AN (Focker et al., 2011; Hebebrand & Bulik, 2011). Weight restoration increases leptin levels comparable to those of healthy participants (Bosy-Westphal et al., 2005; Escobar et al., 2000; Haas et al., 2005; Hebebrand et al., 1997). During the weight gain phase, participants who had a greater increase in leptin early in treatment subsequently had lower weight gain later in treatment (Haas et al., 2005). The authors attribute this to participants with higher leptin levels having decreased their intake later in treatment, but the relationship between leptin and longer-term outcomes is unknown. While leptin levels following weight restoration in AN have had mixed results in relation to 1-year weight outcome in adolescent patients (Holtkamp et al., 2004; Seitz et al., 2016), it is unknown whether or not leptin levels in adults would have a predictive value for long-term weight maintenance outcome. With this, the goal for the current study was to determine if fasting leptin after weight restoration in adult patients with AN is associated with weight maintenance outcome in the year following discharge from an inpatient treatment program.

2 |. METHODS

The study protocols were approved by the institutional review board of the New York State Psychiatric Institute-Columbia University Irving Medical Center (NYSPI/CUIMC). Study was described to the participants and written informed consent was obtained. This study is posted on ClinicalTrials.gov NCT00271921.

2.1 |. Participants and clinical treatment

Participants were 41 women with AN between the ages of 18 and 44 years from a pooled cohort of two previous studies examining longitudinal body composition during treatment of AN (Bodell & Mayer, 2011; Mayer et al., 2007). Both subtypes of AN were included: restricting (AN-R, N = 15) and binge-purge (AN-BP, N = 26). Diagnosis of AN including the amenorrhea criterion was confirmed by Structured Clinical Interview for DSM-IV-TR (Spitzer, Gibbon, & Williams, 1995) during admission to the inpatient unit. Patients were excluded from the study if they met criteria for an additional Axis I disorder other than Major Depression, were on medications including oral contraceptives, had a history of suicide or injurious behavior within 6 months of the study, or had significant current or past medical histories.

Patients were admitted to the Eating Disorders Service of the General Clinical Research Unit (GCRU) at the NYSPI/CUMC and received inpatient treatment for eating disorders, detailed in prior studies (Haynos, Snipes, Guarda, Mayer, & Attia, 2016; Kim, Hildebrandt, & Mayer, 2019). Briefly, treatment aims to normalize weight and eating behaviors until the patient can be weight-restored to at least 90% ideal body weight (90% IBW) as outlined in the 1959 Metropolitan Life Tables (approximately equivalent to BMI of 19.5 kg/m2) (Harrison, 1985). After a few weeks of stability following weight restoration, patients were discharged and followed as outpatients for 1 year.

2.2 |. Timing of data collection

After maintaining the attained goal weight for 2 to 4 weeks as inpatient, the predischarge measurements were collected (Mayer et al., 2007). Upon the completion of the inpatient program, patients were discharged to treatment in the community. Participants were contacted monthly by phone, and in-person evaluations were conducted every 3 months for up to 1 year following hospital discharge. Data from participants who were followed for at least 12 months following discharge were included in the analysis.

2.3 |. Clinical outcome

Treatment outcome was defined using the modified Morgan-Russell criteria (Mayer et al., 2007). Specifically, outcome was dichotomized into two groups using the BMI criterion of the Morgan-Russell criteria. Patients were categorized as having had successful weight maintenance if they had maintained a BMI of at least 18.5 kg/m2 for the 8 weeks leading up to the 1-year follow-up (“maintenance” group), and if they had not, they were categorized as “relapse” outcome group to indicate weight relapse.

2.4 |. Biological parameters

Predischarge testing included height measured to the nearest millimeter using a wall-mounted stadiometer, weight measured to the nearest quarter pound on a beam balance scale (Detecto, Madison, WI), fasting blood sampling for leptin, and consumption of a standardized breakfast (~300 kcal) followed by whole-body body Magnetic Resonance Imaging (MRI) for body composition. BMI was calculated with body weight in kg divided by height in meters squared. Body composition was analyzed through MR imaging at the Program for Imaging in Cognitive Sciences of CUMC and at the St. Luke’s–Roosevelt Hospital through the New York Obesity Research Center using the model 1.5T Twin Speed MRI (General Electric, Milwaukee, WI). Cross sectional images were analyzed by three trained observers with VECT image analysis software (Slice-O-Matic, Montreal, Canada) as previously reported (Mayer et al., 2009).

Venous blood (~10 cc) was drawn from participants after a minimum 8-h overnight fast and before breakfast. Blood samples were processed for plasma at the Core Laboratory of the Irving Institute for Clinical and Translational Research/CTSA using standard techniques. Plasma was stored at −80 °C until subsequent analysis. Fasting leptin levels were assayed using a radioimmunoassay (RIA) kit from Diagnostic Systems Labs (Webster, TX; intra- and interassay coefficients of variation were 2.5% and 3.6%, respectively). Lower limit of detection for leptin was 0.2 ng/ml and all samples were above the lower limits of normal.

2.5 |. Statistical analysis

Significance level was set at .05 and tests were two-tailed unless otherwise indicated. Clinical variables are described as mean ± standard deviation (SD). Intra-group data distribution was tested using the D’Agostino & Pearson Omnibus and Shapiro–Wilk normality tests, and if there was evidence of sampling from a non-Gaussian distribution, nonparametric tests were selected. Group means were compared between outcome groups using the Student’s t-test or the nonparametric Mann–Whitney test. Effect sizes were calculated using the Hedges’ g. Body fat percentage was calculated by fat mass as percentage of weight. To account for known intra- and intersubject variabilities and to address the uneven distribution of hormone data, fasting leptin measures were analyzed after logarithmic transformation.

Logistic regression models were constructed, and models are described with significance testing, exponentiated parameter estimates, and 95%-confidence intervals (pWald, Exp(β) and 95%CI) based on the Wald χ2-test for parameter effects. Overall significance testing for regression models were done using the likelihood ratio (LR) tests compared to the null model (pLR). Improvements in model fit by adding a predictor term in models as covariates was assessed by analysis of variance with prior model(s) with fewer predictor terms and the null model, by assessing for greater reduction in residual deviance, with addition of terms aimed to minimize deviance residuals.

Correlation analyses were conducted using the Spearman’s rank-order method. Given significant correlations between certain variables, regression models with correlated predictors were assessed for collinearity using variance inflation factors (VIF), and all terms in the models presented had low VIF unless otherwise stated. Model with leptin and fat mass had high multicollinearity, and since leptin is produced by adipocytes, fat-adjusted leptin was used as a single term, followed by log-transformation. Analyses were performed with R for Mac OS X, version 4.0.0 GUI 1.71 (http://www.R-project.org) and Prism 5.0a for Mac OS X (GraphPad Software, Inc.).

3 |. RESULTS

The BMI of patients following acute weight restoration and prior to discharge ranged between 19.2 and 22.8 kg/m2. At 1-year follow-up, 20 (48.8%) of 41 patients maintained a BMI of at least 18.5 kg/m2. This “maintenance” outcome group had significantly higher leptin compared to the “relapse” group (log-transformed, p = .010, Hedges’ g = 0.821). Consistent with prior studies, predischarge body fat percentage (BF%) was also higher in patients who maintained weight. Other clinical variables were not significantly different between groups, including age, duration of illness, duration of amenorrhea, return of menses or amenorrheic status, and predischarge BMI (Table 1).

TABLE 1.

Clinical characteristics of the study cohort shortly following weight restoration

All Maintenance Relapse p Hedges’ g
N 41 20 21 — —
Age, years 25.0 ± 5.3 23.8 ± 4.5 26.1 ± 6.0 .213 −0.418
Subtype, %restricting 36.6 40.0 33.3 .751a —
Duration of illness, years 7.4 ± 6.1 5.6 ± 2.5 9.1 ± 7.8 .154 −0.591
Duration of amenorrhea, years 26.5 ± 31.5 21.4 ± 26.8 34.9 ± 38.5 .244 −0.410
Menses, %return of menses 36.6 45.0 28.6 .341a —
Body mass index (BMI), kg/m2 20.4 ± 0.6 20.4 ± 0.5 20.3 ± 0.8 .201 0.223
Body fat percentage (BF%) 25.3 ± 5.0 27.5 ± 4.4 23.1 ± 4.6 .004 0.944
Leptin, ng/ml 10.1 ± 11.7 13.8 ± 14.7 6.6 ± 6.5 .016 0.625
Log(leptin) 1.8 ± 1.1 2.2 ± 0.9 1.4 ± 1.1 .010 0.821
Open in a new tab

Note: Data are shown together and grouped by binary outcome groups “maintenance” and “relapse” where weight maintenance is defined as 1-year follow-up BMI of at least 18.5 kg/m2. Values are presented as mean ± SD. Except as indicated belowa, statistical comparisons between outcome groups are shown as p-values, using unpaired parametric Student’s t-test or nonparametric Mann–Whitney methods, depending on the normality of the sample distribution. Statistically significant results are in bold (p < .05). Sample N values correspond to that shown in the top row, except for the duration of amenorrhea, where N = 21, with N = 13 in the maintenance group and N = 8 in relapse group. Log(leptin) refers to log-transformed leptin values. Menses indicates status of return of menses vs. persistently amenorrheic state after weight restoration.

a

Proportions were tested using the Fisher’s exact test for different proportion in the two outcome groups of subtypes of AN (Fisher’s p = .751, OR = 1.324, 95%CI = [0.311, 5.767]) and menstrual status (Fisher’s p = .341, OR = 0.498, 95%CI = [0.109, 2.121]).

Logistic regression analysis was performed to identify characteristics associated with weight maintenance at 1 year. Each parameter was used alone as an independent predictor variable or together with age and BMI as covariates to account for known associations between higher age (Berends et al., 2016) and lower BMI (Aguera et al., 2015; El Ghoch et al., 2016; Kaplan et al., 2009) with relapse. Specific clinical variables were selected for inclusion in the logistic regression analysis for the following objectives: (a) to test fat-derived hormone leptin as a predictor of outcome, given the association between body fat and outcome and (b) to confirm prior reports of association with weight maintenance for variables age, duration of illness, BMI, and BF% (Berends et al., 2018). Both higher leptin levels and BF% significantly predicted weight maintenance individually or after adjustment for age and BMI (Table 2). Other predischarge variables were not associated with outcome.

TABLE 2.

Binary logistic regression modeling using predischarge parameters as predictors for 1-year outcome

p Wald Exp(β) 95%CI
Independent predictors
Age, years .188 0.915 [0.788, 1.034]
Duration of illness, years .080 0.843 [0.676, 0.987]
Body mass index (BMI), kg/m2 .466 1.460 [0.054, 4.495]
Body fat percentage (BF%) .010 1.260 [1.076, 1.540]
Log(leptin) .021 2.566 [1.255, 6.379]
Log(leptin)fat adj .029 2.744 [1.212, 7.625]
Model with age and BMI
Body fat percentage (BF%) .009 1.314 [1.096, 1.668]
Log(leptin) .024 2.909 [1.293, 8.536]
Log(leptin)fat adj .034 2.962 [1.211, 9.317]
Open in a new tab

Note: Predischarge parameters were used as independent predictors in constructing binary logistic regression models for the binary 1-year outcomes as response variable. Each parameter was used as the single predictor variable alone first, with significant predictors retested with age and BMI as covariates. Significance testing was conducted using Wald χ2-test for coefficients of independent predictor parameters in generalized linear models with binomial response and link logit, as well as likelihood ratio tests based on scaled deviances. Significant p-values are in bold (p < .05). The odds ratio for the predictors calculated by exponentiating the coefficient estimate, Exp(β) and confidence interval (95%CI) for the Exp(β) obtained by exponentiating the confidence interval boundaries for the estimate effect of the variable in the logit scale. Log(leptin), log-transformed leptin; log(leptin)fat adj, log-transformed fat-adjusted leptin. N = 41 for all variables except for duration of amenorrhea, N = 21.

We asked if the predischarge parameters were associated with each other and found significant positive associations between age and duration of illness (rho = 0.557, p = 1.565 × 10−4), BMI and BF% (rho = 0.370, p = .017), log-leptin and BMI (rho = 0.401, p = .009), and log-leptin and BF% (rho = 0.793, p = 6.584 × 10−10). All other pairs of parameters had no significant associations (Supplementary Figure 1). With a highly significant positive relationship between leptin and BF%, where both parameters predicted outcome, we asked if the association between leptin and outcome is a secondary effect from the relationship between BF% and outcome, given that leptin is produced by adipocytes. Indeed, the logistic regression model with both leptin and BF% as covariates indicated a poor fit with significant collinearity (VIF > 10). Therefore, we used the log-transformation of leptin adjusted for fat mass as a predictor variable, and this was significantly associated with outcome, including in models adjusted for age and BMI (pWald = .034, pLR = .040).

Leptin has been associated with reproductive hormone status and return of menses. Unfortunately, only 15 women had return of menses prior to hospital discharge, and the resumption of menses was not associated with leptin levels (rho = 0.086, p = .595). Of them, nine were able to maintain weight and six lost weight by the end of the year. While these differences were not statistically significant (Fisher’s p = .341, OR = 0.498), this sample size is too small to meaningfully explore the relationship between leptin and menses.

4 |. DISCUSSION

In the present work, we show that higher levels of leptin measured shortly after weight restoration were associated with sustained weight maintenance in the year following hospital discharge. As leptin is a hormone produced by fat cells and known to be strongly correlated with fat mass (Lear et al., 1999), this finding is consistent with previous work on the association between weight maintenance and higher predischarge body fat (Bodell & Mayer, 2011; Mayer et al., 2007). As expected, leptin and body fat are robustly associated, which could explain why they were independently associated with outcome in separate regression models, yet the model with both terms as covariates failed due collinearity. Most importantly, leptin adjusted for body fat was independently associated with the weight outcome.

Leptin as a putative predictor of weight maintenance outcome has never been published in adult patient population with AN, and our study is the first to report this finding. In our previous study (Mayer et al., 2007), the average predischarge leptin in participants with successful 1-year weight maintenance was higher than that of weight-relapsed participants, but it did not reach statistical significance, likely reflecting a lack of sufficient power in trying to detect differences for a parameter with high variability using a small sample. The current study utilizes a larger sample size, which could account for the differences.

In adolescent patients, two studies have examined this, and they have published results conflicting with each other and with our findings of adult patients. Holtkamp et al. (Holtkamp et al., 2004) reported that higher leptin levels are associated with weight relapse at 2-month and 1-year follow-up, and Seitz et al. (Seitz et al., 2016) reported a lack of association between leptin and 1-year weight maintenance. The discrepancy between our findings and these studies could be explained by a higher posttreatment BMI and body fat of participants in our study, group differences between adolescents and adults (Misra & Klibanski, 2014), and possibly due to different proportions of AN subtypes (Supplementary Table 1). Regarding the latter, the majority of our cohort is AN-BP, compared to the adolescent studies that are mostly AN-R, and subtype contribution to differential weight maintenance and relapse has been suggested (Carter et al., 2012). Though interesting, the relationship between subtype, leptin, and outcome is difficult to assess meaningfully with our sample size, and a larger study powered to examine the biological differences would be informative.

The mechanism of action underlying the association between leptin and outcome is likely complex. Here, we consider the role of leptin in the feeding circuitry, psychological symptoms, and neuroendocrine recovery. Leptin functions in both the homeostatic and the hedonic feeding neurocircuitry in the brain (Tang-Christensen, Havel, Jacobs, Larsen, & Cameron, 1999; Zhang et al., 1994). The discovery of leptin generated much excitement for it to be a potential “satiety hormone”, but in AN, the near-deficient levels of leptin in the underweight patients does not lead to an increase in feeding behaviors. Rather, the underweight patients appear to have significantly increased self-report satiety scores despite the associated reduction in leptin (Haas et al., 2005) and it is unclear how and if leptin affects feeding in AN. We consider a different possibility. Fasting acutely decreases serum leptin (Kolaczynski et al., 1996; Weigle et al., 1997), and binge-eating behavior can decrease leptin secretion (Eddy et al., 2015; Taylor, Hubbard, & Anderson, 1999). Moreover, in underweight AN, it has been reported that leptin had a significant inverse correlation with various eating disorder symptoms including eating concern, dietary restraint, and the number of binge and purge episodes per week (Eddy et al., 2015). Taken together, it is possible that rather than leptin playing an active role in the regulation of eating behavior in AN, leptin levels may reflect the effect of current disordered eating behaviors superimposed on the chronic leptin-fat energy balance information. Thus, the reduced leptin levels may have captured persistently disordered eating behaviors, which could account for its association with relapse.

Leptin may have an impact on weight maintenance through its effects on psychological symptoms, including mood. Leptin has been shown to have antidepressant properties in humans and animal models of mood disorders (Licinio, Negrao, & Wong, 2014; Lu, Kim, Frazer, & Zhang, 2006), and a study reported that decreased levels of leptin in women with AN were associated with increased symptoms of depression, anxiety, and perceived stress (Lawson et al., 2012). Thus, elevated leptin may contribute to a positive mood regulatory effect that promotes long-term recover in our patient population, which are often seen with improved endocrine health over time.

Multiple perturbations in the endocrine system are observed in the patients with AN, which normalize partly after weight restoration followed by slower progressive recovery during weight maintenance (Miller, 2011; Misra & Klibanski, 2014). Leptin has been linked to the reversal of various metabolic and endocrine abnormalities (Park & Ahima, 2015), raising the possibility that leptin may play a role in weight maintenance in AN through the restoration of the neuroendocrine balance and energy homeostasis (Dalamaga et al., 2013; Hebebrand, Muller, Holtkamp, & Herpertz-Dahlmann, 2007). For instance, in our cohort, there are no significant differences in the proportion of patients with recovery of menses in each outcome group, and it is possible that higher leptin contributes progressively to neuroendocrine recovery during the maintenance phase. Future studies with larger sample size might focus on whether the higher predischarge leptin could predict a faster menstrual recovery.

The strengths of this study include a broad age range, duration of follow-up, and the use of a sophisticated body composition assessment. The weaknesses of this study include the retrospective study design and the limitations in our BMI-based definition of 1-year outcome categories. Our outcome categories were defined with a narrow focus on weight maintenance, but future studies with larger sample sizes could build in additional complexities in the definition of 1-year recovery, and it would be clinically important to determine whether predischarge leptin and body fat are still associated with 1-year outcome defined in a more complex manner. For instance, weight-restored patients who are persistently amenorrhoeic and not fully recovered from AN are known to have lower levels of circulating leptin independent of BMI and body fat (Djurovic et al., 2004), and monitoring of leptin levels could be important during clinical treatment during weight maintenance. In addition, while we characterize the weight maintenance and relapse groups as similar in age, duration of illness, and duration of amenorrhea (Table 1), we acknowledge that there is a possibility that trends of higher age, longer duration of illness and amenorrhea in the relapse group could reflect a true difference given the moderate effect sizes, but we are limited in power given the large variances relative to available observations. Lastly, we used a single measure of fasting leptin, which does not account for the intra- and interindividual differences in the diurnal variation of leptin levels, as well as the different ratios of free- and bound-forms of circulating leptin. However, the fact that we were able to capture the differences in leptin between the outcome groups despite the unmeasured variability together with large effect sizes, suggest that differences in leptin are quite significant.

In summary, in recently weight-recovered women with AN, both higher absolute and body fat-adjusted leptin levels were associated with weight maintenance outcome at 1-year after discharge. While this may be largely driven by body composition, it may also relate to leptin’s complex action as a signal of nutritional status, reflecting both fat storage and relative energy imbalance from persistently disordered eating behaviors. In addition, pleiotropic roles of leptin in restoring multiple endocrine axes, as well as in positive emotion regulation, could suggest a more active role of leptin in promoting recovery in the long run. In fact, it has been suggested that leptin replacement may be a treatment approach for AN given the likely benefit in normalizing leptin during recovery (Hebebrand et al., 2019), and while this is a controversial topic due to potential adverse consequences of leptin replacement therapy, there may be a value in monitoring serum leptin levels during and after weight restoration treatment. Body composition assessment and posttreatment leptin measurements may be informative in identifying patients at higher risk for relapse, and who might, therefore, benefit from an enhanced, targeted relapseprevention effort. Relatedly, studying factors during inpatient treatment that might contribute to variable leptin levels in individual patients could point to aspects of treatment that could be further optimized. Future studies assessing the effect of leptin on other aspects of clinical outcome in weight-restored patients with AN are warranted to better understand leptin as a potential biomarker of outcome in AN.

Supplementary Material

supplementary material

ACKNOWLEDGMENTS

We thank the patients and staff of the Eating Disorders Research Unit of Columbia University Irving Medical Center/New York State Psychiatric Institute without whom this study would not have been possible. This work was supported by the Columbia University Irving Medical Center/New York State Psychiatric Institute Eating Disorders Research Unit, National Institutes of Health grants: R03 DK-066033 and K23 DK-02749; and the NCRR UL1 RR024156 from the National Center for Advancing Translational Sciences. This work was additionally supported by the Columbia University Vagelos College of Physicians and Surgeons Sarah and Arnold P. Friedman Research Award, Leon Levy Foundation Fellowship, and the Icahn School of Medicine at Mount Sinai Physician-Scientist Psychiatry Training Program Support to YK. The content is solely the responsibility of the authors and does not necessarily represent the official views of the funding institutions.

Funding information

Leon Levy Foundation Fellowship; Sarah and Arnold P. Friedman Research Award; National Center for Research Resources, Grant/Award Number: UL1 RR024156; National Institute of Diabetes and Digestive and Kidney Diseases, Grant/Award Numbers: K23 58 DK-02749, R03 DK-066033

Footnotes

CONFLICT OF INTEREST

Dr. Walsh reports receiving royalties and honoraria from McGraw-Hill, Oxford University Press, UpToDate, British Medical Journal, Johns Hopkins Press and Guidepoint Global. Authors have no other conflicts of interest to disclose.

SUPPORTING INFORMATION

Additional supporting information may be found online in the Supporting Information section at the end of this article.

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.

REFERENCES

  1. Aguera Z, Romero X, Arcelus J, Sanchez I, Riesco N, JimenezMurcia S, Gonzalez-Gomez J, Granero R, Custal N, Montserrat-Gil de Bernabe M, Tarrega S, Banos RM, Botella C, de la Torre R, Fernandez-Garcia JC, Fernandez-Real JM, Fruhbeck G, GomezAmbrosi J, Tinahones FJ, Crujeiras AB, Casanueva FF, Menchon JM, & Fernandez-Aranda F. (2015). Changes in body composition in anorexia nervosa: Predictors of recovery and treatment outcome. PLoS One, 10, e0143012. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Arcelus J, Mitchell AJ, Wales J, & Nielsen S. (2011). Mortality rates in patients with anorexia nervosa and other eating disorders. A meta-analysis of 36 studies. Archives of General Psychiatry, 68, 724–731. [DOI] [PubMed] [Google Scholar]
  3. Berends T, Boonstra N, & van Elburg A. (2018). Relapse in anorexia nervosa: A systematic review and meta-analysis. Current Opinion in Psychiatry, 31, 445–455. [DOI] [PubMed] [Google Scholar]
  4. Berends T, van Meijel B, Nugteren W, Deen M, Danner UN, Hoek HW, & van Elburg AA (2016). Rate, timing and predictors of relapse in patients with anorexia nervosa following a relapse prevention program: A cohort study. BMC Psychiatry, 16, 316. [DOI] [PMC free article] [PubMed] [Google Scholar]
  5. Bodell LP, & Mayer LE (2011). Percent body fat is a risk factor for relapse in anorexia nervosa: A replication study. The International Journal of Eating Disorders, 44, 118–123. [DOI] [PMC free article] [PubMed] [Google Scholar]
  6. Bosy-Westphal A, Brabant G, Haas V, Onur S, Paul T, Nutzinger D, … Muller MJ (2005). Determinants of plasma adiponectin levels in patients with anorexia nervosa examined before and after weight gain. European Journal of Nutrition, 44, 355–359. [DOI] [PubMed] [Google Scholar]
  7. Carter JC, Blackmore E, Sutandar-Pinnock K, & Woodside DB (2004). Relapse in anorexia nervosa: A survival analysis. Psychological Medicine, 34, 671–679. [DOI] [PubMed] [Google Scholar]
  8. Carter JC, Mercer-Lynn KB, Norwood SJ, Bewell-Weiss CV, Crosby RD, Woodside DB, & Olmsted MP (2012). A prospective study of predictors of relapse in anorexia nervosa: Implications for relapse prevention. Psychiatry Research, 200, 518–523. [DOI] [PubMed] [Google Scholar]
  9. Chou SH, Chamberland JP, Liu X, Matarese G, Gao C, Stefanakis R, … Mantzoros CS (2011). Leptin is an effective treatment for hypothalamic amenorrhea. Proceedings of the National Academy of Sciences of the United States of America, 108, 6585–6590. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Considine RV, Sinha MK, Heiman ML, Kriauciunas A, Stephens TW, Nyce MR, Ohannesian JP, Marco CC, McKee LJ, Bauer TL, & et al. (1996). Serum immunoreactive-leptin concentrations in normal-weight and obese humans. N Engl J Med, 334, 292–295. [DOI] [PubMed] [Google Scholar]
  11. Dalamaga M, Chou SH, Shields K, Papageorgiou P, Polyzos SA, & Mantzoros CS (2013). Leptin at the intersection of neuroendocrinology and metabolism: Current evidence and therapeutic perspectives. Cell Metabolism, 18, 29–42. [DOI] [PubMed] [Google Scholar]
  12. Djurovic M, Pekic S, Petakov M, Damjanovic S, Doknic M, Dieguez C, … Popovic V. (2004). Gonadotropin response to clomiphene and plasma leptin levels in weight recovered but amenorrhoeic patients with anorexia nervosa. Journal of Endocrinological Investigation, 27, 523–527. [DOI] [PubMed] [Google Scholar]
  13. Eddy KT, Lawson EA, Meade C, Meenaghan E, Horton SE, Misra M, … Miller KK (2015). Appetite regulatory hormones in women with anorexia nervosa: Binge-eating/purging versus restricting type. The Journal of Clinical Psychiatry, 76, 19–24. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. El Ghoch M, Calugi S, Chignola E, Bazzani PV, & Dalle Grave R. (2016). Body mass index, body fat and risk factor of relapse in anorexia nervosa. European Journal of Clinical Nutrition, 70, 194–198. [DOI] [PubMed] [Google Scholar]
  15. Escobar L, Freire JM, Giron JA, Vazquez JM, Pajares M, Ortego J, … Aguilar M. (2000). Plasma levels of insulin and leptin in patients with morbid obesity and anorexia nervosa after weight loss or gain, respectively. Revista Española de Medicina Nuclear, 19, 199–206. [PubMed] [Google Scholar]
  16. Fichter MM, Quadflieg N, Crosby RD, & Koch S. (2017). Long-term outcome of anorexia nervosa: Results from a large clinical longitudinal study. The International Journal of Eating Disorders, 50, 1018–1030. [DOI] [PubMed] [Google Scholar]
  17. Focker M, Timmesfeld N, Scherag S, Buhren K, Langkamp M, Dempfle A, … Hebebrand J. (2011). Screening for anorexia nervosa via measurement of serum leptin levels. Journal of Neural Transmission (Vienna), 118, 571–578. [DOI] [PubMed] [Google Scholar]
  18. Grinspoon S, Gulick T, Askari H, Landt M, Lee K, Anderson E, … Klibanski A. (1996). Serum leptin levels in women with anorexia nervosa. The Journal of Clinical Endocrinology and Metabolism, 81, 3861–3863. [DOI] [PubMed] [Google Scholar]
  19. Haas V, Onur S, Paul T, Nutzinger DO, Bosy-Westphal A, Hauer M, … Muller MJ (2005). Leptin and body weight regulation in patients with anorexia nervosa before and during weight recovery. The American Journal of Clinical Nutrition, 81, 889–896. [DOI] [PubMed] [Google Scholar]
  20. Harrison GG (1985). Height-weight tables. Annals of Internal Medicine, 103, 989–994. [DOI] [PubMed] [Google Scholar]
  21. Haynos AF, Snipes C, Guarda A, Mayer LE, & Attia E. (2016). Comparison of standardized versus individualized caloric prescriptions in the nutritional rehabilitation of inpatients with anorexia nervosa. The International Journal of Eating Disorders, 49, 50–58. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Hebebrand J, Blum WF, Barth N, Coners H, Englaro P, Juul A, … Remschmidt H. (1997). Leptin levels in patients with anorexia nervosa are reduced in the acute stage and elevated upon short-term weight restoration. Molecular Psychiatry, 2, 330–334. [DOI] [PubMed] [Google Scholar]
  23. Hebebrand J, & Bulik CM (2011). Critical appraisal of the provisional DSM-5 criteria for anorexia nervosa and an alternative proposal. The International Journal of Eating Disorders, 44, 665–678. [DOI] [PubMed] [Google Scholar]
  24. Hebebrand J, Milos G, Wabitsch M, Teufel M, Fuhrer D, Buhlmeier J, … Antel J. (2019). Clinical trials required to assess potential benefits and side effects of treatment of patients with anorexia nervosa with recombinant human leptin. Frontiers in Psychology, 10, 769. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Hebebrand J, Muller TD, Holtkamp K, & Herpertz-Dahlmann B. (2007). The role of leptin in anorexia nervosa: Clinical implications. Molecular Psychiatry, 12, 23–35. [DOI] [PubMed] [Google Scholar]
  26. Hebebrand J, van der Heyden J, Devos R, Kopp W, Herpertz S, Remschmidt H, & Herzog W. (1995). Plasma concentrations of obese protein in anorexia nervosa. Lancet, 346, 1624–1625. [DOI] [PubMed] [Google Scholar]
  27. Holtkamp K, Hebebrand J, Mika C, Heer M, Heussen N, & Herpertz-Dahlmann B. (2004). High serum leptin levels subsequent to weight gain predict renewed weight loss in patients with anorexia nervosa. Psychoneuroendocrinology, 29, 791–797. [DOI] [PubMed] [Google Scholar]
  28. Kaplan AS, Walsh BT, Olmsted M, Attia E, Carter JC, Devlin MJ, … Parides M. (2009). The slippery slope: Prediction of successful weight maintenance in anorexia nervosa. Psychological Medicine, 39, 1037–1045. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Keel PK, Dorer DJ, Franko DL, Jackson SC, & Herzog DB (2005). Postremission predictors of relapse in women with eating disorders. The American Journal of Psychiatry, 162, 2263–2268. [DOI] [PubMed] [Google Scholar]
  30. Khalsa SS, Portnoff LC, McCurdy-McKinnon D, & Feusner JD (2017). What happens after treatment? A systematic review of relapse, remission, and recovery in anorexia nervosa. Journal of Eating Disorders, 5, 20. [DOI] [PMC free article] [PubMed] [Google Scholar]
  31. Kim Y, Hildebrandt T, & Mayer LES (2019). Differential glucose metabolism in weight restored women with anorexia nervosa. Psychoneuroendocrinology, 110, 104404. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Kolaczynski JW, Considine RV, Ohannesian J, Marco C, Opentanova I, Nyce MR, … Caro JF (1996). Responses of leptin to short-term fasting and refeeding in humans: A link with ketogenesis but not ketones themselves. Diabetes, 45, 1511–1515. [DOI] [PubMed] [Google Scholar]
  33. Lawson EA, Miller KK, Blum JI, Meenaghan E, Misra M, Eddy KT, … Klibanski A. (2012). Leptin levels are associated with decreased depressive symptoms in women across the weight spectrum, independent of body fat. Clinical Endocrinology, 76, 520–525. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Lear SA, Pauly RP, & Birmingham CL (1999). Body fat, caloric intake, and plasma leptin levels in women with anorexia nervosa. The International Journal of Eating Disorders, 26, 283–288. [DOI] [PubMed] [Google Scholar]
  35. Li D, Chang X, Connolly JJ, Tian L, Liu Y, Bhoj EJ, … Hakonarson H. (2017). A genome-wide association study of anorexia nervosa suggests a risk locus implicated in dysregulated leptin signaling. Scientific Reports, 7, 3847. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Licinio J, Negrao AB, & Wong ML (2014). Plasma leptin concentrations are highly correlated to emotional states throughout the day. Translational Psychiatry, 4, e475. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Lu XY, Kim CS, Frazer A, & Zhang W. (2006). Leptin: a potential novel antidepressant. Proceedings of the National Academy of Sciences of the United States of America, 103, 1593–1598. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. Mantzoros CS, Magkos F, Brinkoetter M, Sienkiewicz E, Dardeno TA, Kim SY, … Koniaris A. (2011). Leptin in human physiology and pathophysiology. American Journal of Physiology. Endocrinology and Metabolism, 301, E567–E584. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Mayer LE, Klein DA, Black E, Attia E, Shen W, Mao X, … Walsh BT (2009). Adipose tissue distribution after weight restoration and weight maintenance in women with anorexia nervosa. The American Journal of Clinical Nutrition, 90, 1132–1137. [DOI] [PMC free article] [PubMed] [Google Scholar]
  40. Mayer LE, Roberto CA, Glasofer DR, Etu SF, Gallagher D, Wang J, … Walsh BT (2007). Does percent body fat predict outcome in anorexia nervosa? The American Journal of Psychiatry, 164, 970–972. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. Miller KK (2011). Endocrine dysregulation in anorexia nervosa update. The Journal of Clinical Endocrinology and Metabolism, 96, 2939–2949. [DOI] [PMC free article] [PubMed] [Google Scholar]
  42. Misra M, & Klibanski A. (2014). Endocrine consequences of anorexia nervosa. The Lancet Diabetes and Endocrinology, 2, 581–592. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Park HK, & Ahima RS (2015). Physiology of leptin: Energy homeostasis, neuroendocrine function and metabolism. Metabolism, 64, 24–34. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Pike KM (1998). Long-term course of anorexia nervosa: Response, relapse, remission, and recovery. Clin Psychol Rev, 18, 447–475. [DOI] [PubMed] [Google Scholar]
  45. Seitz J, Buhren K, Biemann R, Timmesfeld N, Dempfle A, Winter SM, … Focker M. (2016). Leptin levels in patients with anorexia nervosa following day/inpatient treatment do not predict weight 1 year post-referral. European Child & Adolescent Psychiatry, 25, 1019–1025. [DOI] [PubMed] [Google Scholar]
  46. Spitzer R. l., Gibbon M, & Williams JB (1995). Structured clinical interview for Axis I DSM-IV disorders (SCID). Washington, DC: American Psychiatric Press, Inc. [Google Scholar]
  47. Steinglass JE, & Walsh BT (2016). Neurobiological model of the persistence of anorexia nervosa. Journal of Eating Disorders, 4, 19. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Tang-Christensen M, Havel PJ, Jacobs RR, Larsen PJ, & Cameron JL (1999). Central administration of leptin inhibits food intake and activates the sympathetic nervous system in rhesus macaques. The Journal of Clinical Endocrinology and Metabolism, 84, 711–717. [DOI] [PubMed] [Google Scholar]
  49. Taylor AE, Hubbard J, & Anderson EJ (1999). Impact of binge eating on metabolic and leptin dynamics in normal young women. The Journal of Clinical Endocrinology and Metabolism, 84, 428–434. [DOI] [PubMed] [Google Scholar]
  50. Weigle DS, Duell PB, Connor WE, Steiner RA, Soules MR, & Kuijper JL (1997). Effect of fasting, refeeding, and dietary fat restriction on plasma leptin levels. The Journal of Clinical Endocrinology and Metabolism, 82, 561–565. [DOI] [PubMed] [Google Scholar]
  51. Zhang Y, Proenca R, Maffei M, Barone M, Leopold L, & Friedman JM (1994). Positional cloning of the mouse obese gene and its human homologue. Nature, 372, 425–432. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

supplementary material
NIHMS1690573-supplement-supplementary_material.docx (2.6MB, docx)

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.

RESOURCES